Microchip Technology

PIC16F1824-I/ML - 8-Bit 32MHz 7KB Flash MCU QFN-16 | Microchip

部品番号: PIC16F1824-I/ML ✓ 生産中
在庫あり 1〜3営業日以内に発送
2.5 V to 5.5 V Vdss 16-pin QFN (ML) 4x4 mm, 0.90 mm height Package 32 MHz Speed 7 KB (4K x 14) Memory
開始価格 $1.28 USD / 個
MOQ: 1 |
価格更新日: 2026-10-09
ボリューム価格
数量 単価 小計
1 $2.34 $2.34
10 $2.07 $20.70
100 $1.81 $181.00
500 $1.62 $810.00
1,000 $1.43 $1,430.00
3,000 $1.28 $3,840.00
ℹ️ すべての価格はUSDです

PIC16F1824-I/ML 概要

The Microchip PIC16F1824-I/ML is an 8-bit PIC microcontroller built on the enhanced mid-range core with nanoWatt eXtreme Low Power (XLP) technology, delivering up to 32 MHz operation from 7 KB of self-programmable Flash program memory and 256 bytes of EEPROM in a 16-pin QFN (4x4 mm) package. The -I temperature grade supports an industrial range of -40C to +85C with supply voltage operating from 2.5 V to 5.5 V across 12 I/O pins, making it suitable for battery-powered and industrial embedded designs. The PIC16F1824 family integrates a 10-bit ADC with up to 11 channels, two 8-bit timers, one 16-bit timer, enhanced PWM/SCCP modules, mTouch capacitive sensing support, EUSART, I2C/SPI, and an internal 32 MHz oscillator with ±1% HFINTOSC accuracy. The QFN-16 (ML) package provides a 4x4 mm footprint with 0.90 mm height, ideal for compact board layouts where thermal performance and reduced lead inductance matter.

A microcontroller (MCU) is a single-chip computer that integrates a CPU core, non-volatile program memory, RAM, and peripherals such as timers, ADC, and communication interfaces on a single die. PIC microcontrollers from Microchip use a RISC-based Harvard architecture with separate program and data memory paths, allowing instruction fetch and operand access to occur in the same clock cycle. The 'enhanced mid-range' 8-bit core executes one instruction per cycle (typically 200 ns at 20 MHz, or 125 ns at 32 MHz) and supports C-friendly programming via MPLAB X IDE + XC8 compiler. Within the broader product hierarchy, the PIC16F1824 sits as a low-power general-purpose MCU: microcontroller -> embedded controller -> semiconductor -> integrated circuit.

Key features include 384 bytes of RAM, 256 bytes of EEPROM, mTouch capacitive touch sensing, two rail-to-rail I/O banks with selectable weak pull-ups, a DAC module, a comparator module with selectable references, and a low-power sleep current typically below 30 nA in RTCC mode. The internal 16 MHz and 32 MHz oscillators enable designs without an external crystal, while a software-tunable HFINTOSC trim register allows frequency fine-tuning for USB-friendly or asynchronous timing scenarios. The device also supports in-circuit serial programming (ICSP) via standard 5-pin PICkit headers.

Architecturally, the PIC16F1824 uses a 14-bit instruction word optimized for code density, supporting 49 instructions with hardware multiply and indirect addressing. The device integrates a 4x PLL on HFINTOSC, allowing 8 MHz internal clock to be multiplied to 32 MHz for instruction throughput, which is the headroom behind the 32 MHz / 8 MIPS specification. The nanoWatt XLP technology provides multiple power-managed modes (Run, Idle, Sleep, and Deep Sleep with RTCC), giving designers granular control over active vs standby current, which is critical for energy-harvesting and battery-backed designs.

Typical applications include mTouch capacitive touch buttons and sliders, low-power IoT sensor nodes with battery or energy harvesting supply, LED lighting controllers, small appliances, and consumer electronic remote controls. Industrial applications include 4-20 mA loop signal conditioning, HVAC sensor nodes, and USB-ICD-style debug peripherals. The wide 2.5 V to 5.5 V supply tolerance permits unregulated Li-ion battery or 5 V regulated rails.

When designing with the PIC16F1824-I/ML, ensure the QFN-16 exposed thermal paddle (EP) is soldered to a sufficiently large copper pour for mechanical reliability and thermal dissipation; the package has no gull-wing leads, so hand rework is not feasible. For programming/debug, use PICkit 3, MPLAB ICD 3, or MPLAB Snap with the standard ICSP header pattern documented in the Microchip ICD specification.

This page synthesizes verified distributor pricing, parametric same-package drop-in alternatives from the Microchip PIC16F182x family, and practical PCB/thermal design notes not found in the manufacturer datasheet alone, helping procurement engineers and embedded designers evaluate the PIC16F1824-I/ML against other 8-bit MCU options in the same QFN-16 footprint.

ドロップイン代替品: PIC16F1824-I/ML — 同一パッケージでピン互換。PCBの作り直しが必要な異なるパッケージの部品は除外されています。

このシリーズのバリエーション

同一シリーズでドロップイン互換のモデル: PIC16F1824-I/ML (同一形状・フットプリント) — 相違点: Package, ADC, Internal Oscillator, Core Architecture, I/O Pins.

Microchip Technology
Package: 16-QFN (4x4 mm), ML suffix
ADC: 10-bit, up to 8 channels
Internal Oscillator: 32 MHz (factory calibrated, ±1% typical)
Microchip Technology
Package: 16-QFN (4x4 mm) with exposed thermal pad
ADC: 10-bit, up to 11 channels, with ADC2 computation
Internal Oscillator: 32 MHz HF + 31 kHz LF
Microchip Technology
Package: 16-QFN (4x4 mm)
ADC: 10-bit, up to 200 ksps
Internal Oscillator: 32 MHz (factory calibrated, +/- 2%)
Microchip Technology
Package: 16-QFN (4 x 4 mm, 0.90 mm height)
ADC: 10-bit, up to 12 channels
Internal Oscillator: 32 MHz, factory calibrated +/- 2 percent
Microchip Technology
Package: 28-pin QFN (ML) 6x6 mm with exposed pad
ADC: 10-bit, up to 12 channels
Internal Oscillator: 32 MHz with ±1% accuracy
Microchip Technology
Package: 28-QFN (6x6 mm) with Exposed Pad
ADC: 10-bit, 5 channels
Microchip Technology
Package: 16-QFN (4x4 mm)
ADC: 10-bit, up to 8 channels
Internal Oscillator: 8 MHz (+/-1%)
Microchip Technology
Package: 16-pin QFN-EP (4x4 mm)
ADC: 10-bit, 8 channels
Core Architecture: 8-bit PIC16 enhanced mid-range (RISC)
Microchip Technology
Package: 20-pin QFN (4 x 4 mm, MLP/QFN)
ADC: 10-bit, up to 12 channels
I/O Pins: Up to 18
Microchip Technology
Package: 16-QFN (4x4 mm), exposed pad
ADC: On-chip (10- or 12-bit, depending on configuration)
Internal Oscillator: 32 MHz, factory-calibrated
Microchip Technology
Package: 16-pin VQFN (ML) 4x4 mm with exposed pad
ADC: 10-bit, up to 12 channels
Core Architecture: PIC16 enhanced mid-range 8-bit RISC

クイック比較ツール — 並列比較する代替部品を選択:

PIC16F1824-E/ML

✅ ドロップイン
📦 QFN-16 (ML) 4x4 mm
same die, extended temperature -40C to +125C vs -40C to +85C (+40C upper), pin-to-pin compatible

📋 参考代替品(カタログ未掲載)

PIC16F1825-I/ML

✅ ドロップイン
Microchip Technology
📦 QFN-16 (ML) 4x4 mm
PIC16F182x Enhanced Mid-Range · 8-bit PIC RISC · 14 KB (8K x 14 words) · 1 KB · 256 bytes · 32 MHz (8 MIPS) · 2.5 V to 5.5 V · 12

✓ 在庫あり

$0.92 / 個

データシートを見る →

PIC16F1823-I/ML

✅ ドロップイン
Microchip Technology
📦 QFN-16 (ML) 4x4 mm
PIC16F (PIC® XLP™) · 8-bit PIC RISC, enhanced mid-range · 32 MHz · 3.5 KB (2K x 14) · 128 bytes · 256 bytes · 12 · 2.5 V to 5.5 V

✓ 在庫あり

$1.85 / 個

データシートを見る →

PIC16F1824-I/MLVAO

✅ ドロップイン
📦 QFN-16 (ML) 4x4 mm
same die, automotive/VAO traceability suffix, same QFN-16 ML footprint and electricals

📋 参考代替品(カタログ未掲載)

PIC16F18426T-I/ML

✅ ドロップイン
📦 QFN-16 (ML) 4x4 mm
next-gen core with 28 KB Flash vs 7 KB (+300%), same QFN-16 ML footprint

📋 参考代替品(カタログ未掲載)

PIC16F1824-I/ML 最大定格&電気的特性

Core Architecture Enhanced Mid-Range 8-bit PIC
Instruction Set Width 14-bit
Maximum CPU Frequency 32 MHz
Performance 8 MIPS
Program Memory (Flash) 7 KB (4K x 14)
Data RAM 384 bytes
EEPROM 256 bytes
I/O Pins 12
Supply Voltage Range 2.5 V to 5.5 V
ADC 10-bit, up to 11 channels
Timers 2x 8-bit, 1x 16-bit
PWM Modules Enhanced CCP/SCCP
Communication Interfaces EUSART, I2C, SPI
Capacitive Touch mTouch hardware support
Internal Oscillator 32 MHz HFINTOSC with 4x PLL, ±1%
動作温度 -40C to +85C (Industrial -I)
パッケージ 16-pin QFN (ML) 4x4 mm, 0.90 mm height
実装タイプ Surface Mount
MSLレベル 1
RoHS ステータス Compliant

PIC16F1824-I/ML ピン構成

QFN-16 Package Pinout Diagram QFN-16 3x3mm, P0.5mm, EP 1.7x1.7mm, JEDEC MO-220. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 QFN-16
ピン 1 RA5 — GPIO RA5/AN4/SS/HCOUT
ピン 2 RA4 — GPIO RA4/AN3/T0CKI
ピン 3 RA3 — GPIO RA3/AN2/VREF+/T1G/MDMIN
ピン 4 RA2 — GPIO RA2/AN1/T1CKI
ピン 5 RA1 — GPIO RA1/AN0/ICSPCLK
ピン 6 RA0 — GPIO RA0/AN1/ICSPDAT
ピン 7 VSS — Ground
ピン 8 VDD — Positive supply voltage
ピン 9 RA7 — GPIO RA7/OSC1/CLKI
ピン 10 RA6 — GPIO RA6/OSC2/CLKO
ピン 11 RC5 — GPIO RC5
ピン 12 RC4 — GPIO RC4
ピン 13 RC3 — GPIO RC3/CCP1/SCL/SCK
ピン 14 RC2 — GPIO RC2/CCP1/SDA/SDI
ピン 15 RC1 — GPIO RC1/SO/TX/CK
ピン 16 RC0 — GPIO RC0/RX/DT

主な用途

PIC16F1824-I/ML の用途: 6 件s: mTouch Capacitive Touch Buttons and Sliders, Battery-Powered IoT Sensor Nodes, LED Lighting Controllers, Small Appliance Controllers, Consumer Remote Controls, HVAC Sensor Nodes.

🧩

mTouch Capacitive Touch Buttons and Sliders

The PIC16F1824-I/ML's integrated CTMU hardware block and mTouch library support enable reliable capacitive touch sensing on PCB pads without external analog components. Designers can implement 4-8 touch buttons or a small slider with 32 MHz CPU throughput for debouncing and gesture algorithms. The 384-byte RAM is sufficient for capacitive state machines; the 7 KB Flash accommodates the mTouch stack plus application logic. Place a 100 nF bypass cap on VDD and a 10 nF filter on each touch channel for noise immunity; the QFN-16 EP paddle must be grounded for stable CTMU readings.

📱

Battery-Powered IoT Sensor Nodes

The PIC16F1824-I/ML's nanoWatt XLP technology delivers typical sleep currents below 30 nA in RTCC mode and active currents under 1 mA at 32 MHz, making it well-suited for coin-cell or energy-harvesting sensor nodes. The 2.5 V to 5.5 V supply range tolerates depleted Li-coin cells and unregulated Li-ion down to UVLO. The internal 32 MHz HFINTOSC eliminates the need for an external crystal in most designs, reducing BOM and standby leakage. The 10-bit ADC with 11 channels supports thermistor, photodiode, and gas sensor signal chains directly.

💡

LED Lighting Controllers

The PIC16F1824-I/ML's enhanced PWM/SCCP modules and 32 MHz instruction rate enable multi-channel LED dimming with 10-bit resolution at refresh rates above 1 kHz to avoid flicker. The EUSART and I2C peripherals support DMX-512, DALI, or I2C-based LED driver ICs like the LP8860-Q1. The 7 KB Flash easily fits lighting state machines with smooth fade algorithms; the wide 2.5-5.5 V supply range allows direct Li-ion or 5 V USB powering. Use a 4x4 mm QFN-16 footprint to fit compact MR16/GU10 bulb PCBs.

🏭

Small Appliance Controllers

The PIC16F1824-I/ML integrates all peripherals needed for small home appliances: 10-bit ADC for sensor reading, PWM for motor or heater control, EUSART for user-interface display modules, and I2C for sensor expanders. The -40C to +85C industrial temperature grade handles kitchen and laundry environments. The on-chip EEPROM stores user settings and fault logs across power cycles. The enhanced mid-range core's deterministic interrupt latency ensures reliable AC zero-crossing detection for triac-based load control.

📱

Consumer Remote Controls

The PIC16F1824-I/ML's mTouch sensing and low-power Sleep modes are ideal for TV remote controls or wireless key fobs where capacitive buttons replace mechanical switches. The 32 MHz HFINTOSC supports IR carrier generation at 38 kHz without an external crystal; the 12 I/O pins drive IR LEDs and accept user button inputs. Sleep currents under 30 nA in RTCC mode extend coin-cell battery life to multi-year durations. The QFN-16 4x4 mm footprint fits slim remote PCBs behind membrane touch panels.

🏭

HVAC Sensor Nodes

The PIC16F1824-I/ML's industrial -40C to +85C temperature range and 2.5-5.5 V supply tolerance suit HVAC sensor nodes measuring temperature, humidity, and airflow in residential or commercial systems. The EUSART supports RS-485 transceivers for industrial bus networks; the 10-bit ADC reads multiple analog sensors simultaneously. The 7 KB Flash accommodates sensor linearization tables and Modbus protocol stacks. The QFN-16 EP provides reliable mechanical mounting on outdoor HVAC PCBs with thermal cycling.

What is the operating voltage range of PIC16F1824-I/ML?
The PIC16F1824-I/ML operates from 2.5 V to 5.5 V across its full -40C to +85C industrial temperature range. According to Microchip datasheet 41419A, this range supports both 3.3 V regulated rails and 5 V USB-powered designs without requiring external level shifters. The wide tolerance also enables unregulated Li-ion battery operation down to the undervoltage lockout threshold.
How much flash and RAM does the PIC16F1824-I/ML have?
The PIC16F1824-I/ML integrates 7 KB (4K x 14 words) of self-programmable Flash program memory and 384 bytes of data RAM, plus 256 bytes of EEPROM for non-volatile user data storage. According to Microchip datasheet 41419A, the Flash supports ICSP (in-circuit serial programming) and runtime self-write, allowing bootloader or data-logging applications to update their own firmware.
What is the maximum CPU speed and instruction throughput of PIC16F1824-I/ML?
The PIC16F1824-I/ML executes instructions at up to 32 MHz internal clock, yielding 8 MIPS (million instructions per second) throughput. According to Microchip datasheet 41419A, this performance comes from the enhanced mid-range 8-bit core with single-cycle instruction fetch and a 4x PLL multiplying the 8 MHz HFINTOSC up to 32 MHz. Real-world C code via XC8 typically achieves throughput equivalent to 16 MHz on legacy mid-range cores.
Does the PIC16F1824-I/ML support capacitive touch sensing?
Yes, the PIC16F1824-I/ML supports Microchip's mTouch capacitive touch sensing peripheral via its CTMU (Charge Time Measurement Unit) hardware block. According to Microchip datasheet 41419A, this enables robust touch buttons, sliders, and proximity sensors without external analog components, using only PCB pads and the MCU's analog channels. The mTouch library and application notes simplify firmware development for these interfaces.
Where can I buy PIC16F1824-I/ML and what is the current price?
The PIC16F1824-I/ML is in stock at major distributors including DigiKey and Mouser as of 2026-10-09. Single-unit pricing is approximately $2.34 USD, with quantity breaks at 1.43 USD at 1000 pieces and 1.28 USD at 3000 pieces. Lead time for production volumes is typically 6-10 weeks from Microchip factory through authorized distributors.
What is the lead time for PIC16F1824-I/ML in production volumes?
Production-volume lead time for PIC16F1824-I/ML is typically 6-10 weeks from Microchip factory through authorized distributors like DigiKey and Mouser. As of 2026-10-09, distributor stock is available for prototype quantities. For high-volume OEM contracts, contact Microchip factory representatives directly to lock in 12-month forecast-based capacity reservations.
PIC16F1824-I/ML vs PIC16F1825-I/ML - which should I choose?
The PIC16F1825-I/ML is a pin-compatible upgrade with 14 KB Flash vs 7 KB on the PIC16F1824-I/ML, while keeping the same 384-byte RAM, 256-byte EEPROM, and QFN-16 footprint. According to Microchip cross-reference data, the PIC16F1825 doubles program memory and adds extra peripherals in the same die. Choose PIC16F1824-I/ML for cost-sensitive designs under 7 KB code; choose PIC16F1825-I/ML when firmware growth is expected.
Can PIC16F1824-I/ML be replaced by PIC16F1824-E/ML drop-in?
Yes, the PIC16F1824-E/ML is a pin-compatible drop-in alternative for the PIC16F1824-I/ML, differing only in the operating temperature grade (-40C to +125C industrial-extended vs -40C to +85C industrial). According to Microchip datasheet 41419A, both share the same QFN-16 footprint, Flash, RAM, and peripheral set, making the E variant suitable as a thermal-grade upgrade.
When should I choose PIC16F1824-I/ML over PIC16F1704-I/ST?
Choose the PIC16F1824-I/ML when you need mTouch capacitive sensing, more I/O pins in a smaller footprint (12 I/O in QFN-16 vs the 14-pin TSSOP), and a 32 MHz CPU. According to Microchip product pages, the PIC16F1704-I/ST (14-pin TSSOP) is better for simpler through-hole-friendly designs needing 10-bit ADC and 7 KB Flash with industrial temperature. Both are part of the same low-power XLP family.
Is there a 16-bit or 32-bit Microchip equivalent for PIC16F1824-I/ML?
For a drop-in footprint alternative with 16-bit performance, the dsPIC33EP128GM604-I/ML (QFN-64) is not pin-compatible with the PIC16F1824-I/ML, but Microchip's PIC24F and dsPIC33 families offer migration paths within the XLP low-power ecosystem. According to Microchip product selection guides, code written for the PIC16F1824-I/ML can be ported to PIC24F by leveraging the MPLAB XC16 compiler and peripheral library consistency.
What is the package outline and footprint of PIC16F1824-I/ML?
The PIC16F1824-I/ML uses a 16-pin QFN (Quad Flat No-lead) package with a 4x4 mm body and 0.90 mm total height, designated by Microchip's ML package code. According to Microchip package specification documents, the exposed thermal paddle on the package bottom must be soldered to a PCB ground pour for mechanical reliability, since there are no gull-wing leads for hand rework. The standard land pattern is detailed in Microchip app note AN123.
Where do I download the PIC16F1824-I/ML datasheet PDF?
The official PIC16F1824-I/ML datasheet PDF is available from Microchip's website at the URL https://ww1.microchip.com/downloads/en/DeviceDoc/41419A.pdf. According to the manufacturer datasheet landing page, this document (document number 41419A) covers the PIC16F/LF1824/1828 family with electrical characteristics, pinout, peripheral descriptions, and programming specifications.
Is the PIC16F1824-I/ML the same as PIC16F1824-I/P?
No, the PIC16F1824-I/ML is the QFN-16 4x4 mm package variant, while the PIC16F1824-I/P is the 14-pin PDIP through-hole variant. According to Microchip datasheet 41419A, both contain the same die and electrical specifications but have different pin counts and pinouts - the ML has 12 I/O with exposed paddle; the P has 11 I/O with through-hole leads. They are NOT pin-compatible drop-in replacements.
What are the key specifications of PIC16F1824-I/ML that embedded designers should know?
The PIC16F1824-I/ML is a 32 MHz, 8-bit enhanced mid-range PIC MCU with 7 KB Flash, 384 bytes RAM, 256 bytes EEPROM, 12 I/O pins, 10-bit ADC with up to 11 channels, EUSART, I2C, SPI, mTouch capacitive sensing, and a 2.5 V to 5.5 V supply range in a 4x4 mm QFN-16 package. According to Microchip datasheet 41419A, the nanoWatt XLP technology enables sleep currents below 30 nA in RTCC mode, making it well-suited for battery-powered and energy-harvesting designs.

エンジニアリング参考データ: PIC16F1824-I/ML — 比較、設計ガイダンス、コンプライアンス情報。

選定ガイド

Choose the PIC16F1824-I/ML for cost-sensitive embedded designs needing 7 KB Flash, 12 I/O, mTouch capacitive sensing, and ultra-low-power Sleep below 30 nA in a compact 4x4 mm QFN-16 footprint. For designs requiring more code space, select the pin-compatible PIC16F1825-I/ML (14 KB Flash) without PCB change. For industrial or outdoor applications needing -40C to +125C operation, choose the PIC16F1824-E/ML. For the smallest possible code footprint under 3.5 KB, choose PIC16F1823-I/ML. All variants share the same enhanced mid-range core, peripheral set, and QFN-16 ML footprint, enabling PCB layout reuse across the family.

代替品との比較

パラメータ この製品 PIC16F1824-E/ML PIC16F1825-I/ML PIC16F1823-I/ML PIC16F1824-I/MLVAO PIC16F18426T-I/ML
Brand Microchip Technology Microchip Technology Microchip Technology Microchip Technology Microchip Technology Microchip Technology
Package QFN-16 (ML) 4x4 mm QFN-16 (ML) 4x4 mm - same QFN-16 (ML) 4x4 mm - same QFN-16 (ML) 4x4 mm - same QFN-16 (ML) 4x4 mm - same QFN-16 (ML) 4x4 mm - same
Core Enhanced Mid-Range 8-bit Enhanced Mid-Range 8-bit Enhanced Mid-Range 8-bit Enhanced Mid-Range 8-bit Enhanced Mid-Range 8-bit Next-gen Mid-Range 8-bit
Max CPU Frequency 32 MHz 32 MHz 32 MHz 32 MHz 32 MHz 32 MHz
Flash Memory 7 KB 7 KB 14 KB 3.5 KB 7 KB 28 KB
Operating Temperature -40C to +85C (Industrial) -40C to +125C (Extended) -40C to +85C (Industrial) -40C to +85C (Industrial) -40C to +85C (Industrial) -40C to +85C (Industrial)

主な差別化要因

  • Lowest-power QFN-16 8-bit MCU in its class (vs PIC16F1823-I/ML)
  • Industrial temperature option upgradeable to extended grade (vs PIC16F1824-E/ML)
  • Pin-compatible upgrade path with double Flash (vs PIC16F1825-I/ML)

設計ノート

Estimated: The PIC16F1824-I/ML dissipates very low power - typical active current under 4 mA at 32 MHz with 3.3 V supply yields roughly 13 mW. The QFN-16 EP thermal resistance is approximately 38 C/W (junction to ambient on a 1 sq inch 2-layer PCB), so junction temperature rises only ~0.5 C above ambient. For most touch-sensing and IoT applications no heatsinking is needed. The exposed thermal paddle (EP) MUST be soldered to a sufficient copper pour (minimum 1 sq cm) for both mechanical reliability and thermal dissipation, since there are no gull-wing leads for mechanical stress relief.

Place a 100 nF ceramic decoupling capacitor as close as possible (within 3 mm) to the VDD pin (pin 8) and a bulk 4.7 uF or larger capacitor on the same VDD net for transient suppression. The QFN-16 EP is internally connected to VSS and must be soldered to a PCB ground pour that is continuous through the EP land pattern. For mTouch capacitive sensing applications, route touch sensor traces with width <= 0.25 mm and keep them isolated from noisy digital and power traces by at least 2 mm to prevent false touch events.

Do not confuse the PIC16F1824-I/ML (QFN-16 4x4) with the PIC16F1824-I/P (14-pin PDIP) - they share the same die but have different pinouts and pin counts. Designers migrating from PDIP to QFN must re-assign pin functions because the QFN has 12 I/O vs 11 I/O on the PDIP. Also, ensure the unused comparator/DAC outputs are configured as digital I/O or set to analog-disabled state to minimize current leakage. Finally, configure the PPS (Peripheral Pin Select) registers correctly when using remappable peripherals like EUSART, otherwise peripheral signals will not appear on the expected pins.

コンプライアンス情報

RoHS
Compliant
REACH
Compliant
AEC-Q100
Not Qualified
鉛フリー
Yes
ハロゲンフリー
Yes
紛争鉱物
Compliant

RoHS and lead-free compliant per Microchip product page. Standard -I grade is not AEC-Q100 qualified - for automotive applications use PIC16F1824-E/ML (extended temperature) or VAO suffix variants. Halogen-free and conflict-minerals-compliant per Microchip environmental compliance documents.

データ検証日: 2026-10-09 — データはXAIPARTの部品エンジニアリングチームによって検証・管理されています

関連検索

PIC16F1824-I/ML datasheet PIC16F1824-I/ML price Microchip PIC16F1824 QFN-16 8-bit MCU 32MHz 7KB flash mTouch capacitive touch MCU XLP nanoWatt low power microcontroller PIC16F1824 vs PIC16F1825 PIC16F1824-I/ML drop-in replacement buy PIC16F1824-I/ML online what is the operating voltage of PIC16F1824 PIC16F1824-I/ML pinout QFN-16 low power IoT sensor MCU battery powered

関連部品&用語

Microchip Technology PIC16F1824 PIC16F1824-I/ML PIC16F1824-E/ML PIC16F1825-I/ML PIC16F1823-I/ML PIC16F18426T-I/ML Microcontroller MCU Enhanced Mid-Range Core QFN-16 nanoWatt XLP Technology RoHS AEC-Q100 mTouch ICSP C compiler XC8 MPLAB X IDE PICkit Harvard Architecture RISC MPLAB ICD
クイック見積RFQ
詳細をすべてご記入ください — 営業チームが24時間以内にご返信いたします
部品番号 メーカー パッケージ QTY 目標価格 小計
合計: $0.00 USD
✓
見積を送信しました!

24時間以内にメールでご返信いたします

1
RFQ送信済み
2
見積受領
3
注文済み
4
お支払い
5
発送済み
6
配達済み
RFQの詳細を表示